PROJECT SUMMARY/ABSTRACT Among older adults, falls are a critical issue reducing quality of life and increasing health care costs. Older adults rely on sensory information, such as vibration and light touch-pressure, for balance control. However, somatosensory information declines with age and disease. Fall risk doubles in older individuals with loss of foot sensation8 and increases 12-58% in persons with diabetes prone to sensory related peripheral neuropathy (PN).9 Although studies have linked loss of sensation with fall risk and reduced balance control, the sensory thresholds, such as pressure sensation threshold (PST) and vibration pressure threshold (VPT) for age- and disease-related impaired balance control are not well-established. Progressive pathological changes associated with age- and disease-related somatosensation loss occur on a continuum.9, 10 PST and VPT are easy to measure and commonly used to screen for PN. However, current cut- points for PN classification are based on loss of protective sensation needed for skin integrity. The point at which loss of sensation increases the likelihood of falling has not been established. Use of cut points for PN based on risk of skin ulceration alone may result in the delayed identification of individuals at risk for falls. Identification of falls-specific somatosensory screening tests may improve early detection of fall risk, which will improve quality of life and lower medical expenses for older adults. The long-term goal of this research is to identify fall specific cut-points for PST and VPT and to examine the predictive validity of fall-specific PST and VPT cut-points in combination with the Stopping Elderly Accidents, Deaths and Injury (STEADI) algorithm. In this study, we will establish normative data for PST and VPT in healthy adults (3 groups: 50 young adults aged 25-45 years, 50 adults aged 46-65 years, 50 older adults aged 65-85 years), people with PN (n=50) and older adults with history of fall (n=50). PST and VPT at 10 sites of the feet will be measured using Semmes-Weinstein monofilaments and a biothesiometer, respectively. Older adults with fall history and individuals with PN will be recruited to identify PST and VPT cut-points associated with impaired balance control using multinomial/binary logistic and discriminant function (DFA) approaches and stepwise logistic discriminant analysis (SWLDA). Classification of balance control and fall risk will be based on computerized posturography (sensory organization testing). The effect of updated somatosensory thresholds on the predictive accuracy of the STEADI screening algorithm will be examined for both retrospective and prospective fall. Somatosensory cut-points and clinical fall-risk measures will be monitored at 3-time points over 12 months, allowing longitudinal analysis. Analysis approaches for fall- prediction models will include decision/regression trees to validate cut-points and logistic and DFA methods at each time point and explore the use of generalized logistic models to predict falls over all times.